If I understand it right.. it will be much more costly to host videocentric platform than other types of content. A serious proposal for sharing the burden of hosting will likely be vital - through funding or decentralized storage/processing. Havent heard about that yet.
this post was submitted on 18 Nov 2024
229 points (99.1% liked)
Fediverse
28499 readers
306 users here now
A community to talk about the Fediverse and all it's related services using ActivityPub (Mastodon, Lemmy, KBin, etc).
If you wanted to get help with moderating your own community then head over to !moderators@lemmy.world!
Rules
- Posts must be on topic.
- Be respectful of others.
- Cite the sources used for graphs and other statistics.
- Follow the general Lemmy.world rules.
Learn more at these websites: Join The Fediverse Wiki, Fediverse.info, Wikipedia Page, The Federation Info (Stats), FediDB (Stats), Sub Rehab (Reddit Migration), Search Lemmy
founded 2 years ago
MODERATORS
the bottleneck to peertube seems to be populating and having more viewers, not hosting costs.
The problem with Peertube, IMO, is you can't just "join".
Pick an instance, it's locked down. I get it, hosting others video content is costly and dangerous, but it's also a huge barrier.
It's best use case is self-hosting by established creators funded primarily by patreon subscribers, rather than new creators. But established creators won't abandon the platform with their viewers. So, it's a bit of a jam.
Fediverse platforms also just seem terribly uninterested in supporting creatives' business models, so we get a high minded desert.
Its true, but we do also have a proven base of both users and experienced admins that has proven to be resilient and self-sustaining even in the world where the winds of capitalism or VC money aren't throttling growth.
To me, an alternative model would be deep-linking/embedding 3rd party videos that you own.
For instance, sign up at @server, then:
- Connect your YouTube account
- Post Video 1 to YouTube
- You now have a Video 1 in your federated feed with a YouTube embed
- Connect your Vimeo account
- Post Video 1 to Vimeo (same hash and/or title)
- Your feed entry of Video 1 is updated to support both back end videos
This way your "feed" is federated, but the efficiency of centralized video content is used. If a provider goes down, or you choose to leave, you can re-up to another provider and with the connected account/title matching, your feed could be auto-updated.
That’s a very interesting idea. It might also incentivize creators because it gives them a more stable audience that’s at least a little insured against viewership changes on any single platform due to changes in that platform outside of their control.
not sure about peertube audience size, but afaik it does have a serious subsystem of resource sharing.
Object Storage is relatively cheap, and goes a long way towards affordable hosting. Processing video, on the other hand...
I wonder about this a lot. The little research I did suggested DigitalOcean is footing the bill for the moment (and also for Pixelfed? would love to hear more about this). Google, Facebook, TikTok, etc.. have all managed to throw enough resources at similar products that people expect a level of performance that is very expensive to maintain. There is some serious hardware and distribution issues ($$$$) with trying to host an "instant and endless stream of short form video".
In a counter point though I think large instances like lemmy.world and lemmy.ml have found ways to survive and thrive and the fediverse generally seems to be supported in a very grassroots sort of fashion. Donations, patrons, people who have the hardware and bandwidth sharing what they can for the greater community. Perhaps loops will go the same way.
the videos get automatically deleted after some time.
The very first Youtube video is still up. Any serious competitor is going to need to offer that level of reliability with added benefits to woo users over.
have you considered that a platform for sharing short, vertical videos, developed by a single person, perhaps isn't trying to become a "serious competitor" for Youtube?
Loops isn't competing with youtube. I doubt you would be able to find the first tiktok.
YouTube can't even keep up with the amount of 4k and 8k video being fed into it.
what you're seeing with ads is them trying to reduce their burden.
still waiting for an email about account registration from them.
I got mine within a few minutes.
i am so happy for you. unfortunately i am still waiting.
update: I am in! hooray!
(app looks clean af as well)
Same
Oh nice, the Android app is out!
I can give it a try now
Link to download page within the account settings, for those that have been onboarded already: https://loops.video/dashboard/get-the-app
Edit: it feels very smooth
Yeah the app seems quite polished already around the video viewing experience at least. Wonder how it populates the default feed.
Yeah since there is little content I can’t tell if there is an algorithm or not.
Edit; I tried posting myself to see and the “feed” is a “new” feed of posts by everyone. So no sort of algorithm at all or any customisation available atleast on the ios app, as of now.
How did you get an account?
Signed up on loops.video a couple weeks ago
It's making me login to see that page
I downloaded it without needing an account first (You do need an account to use the actual app), maybe just go on loops.video
v smooth. too smooth. I like it.
Loops hasn’t yet released its source code or enabled federation, but project lead Dan Supernault has stated that he will do both in the near future.
Buying into promises is a gamble dependant on the crediblity and trustworthiness of the person/entity making them
Its always a risk putting your faith into a single person, but Dan is the same guy who built Pixelfed. He's a Fediverse OG, literally the exact opposite of a grifter. Trusting him is a risk for sure, but I can't think of many other developers I'd trust more than him (at least in terms of his intentions and determination... I'm not a skilled dev so I can't speak to whether he actually does good work, but Pixelfed seems incredibly legit given the limited resources he's worked with)
Ah thank you! I didn't realize it was made by Dan. In that case Loops is probably a safer bet than the current alternatives👍
No source code and no federation just makes this a worse app with less liability for bad actions. Not sure why he's pushing this before those are ready. Pixelfed itself isn't exactly finished yet.
At a guess, bc the timing might seem right with all the controversy surrounding Tiktok? Knowing that this is out there may increasing the number of contributors to the codebase.
Also I recall clicking on a GitHub page at some point for a front-end web UI for this content, so while the sourcecode of the app itself may not be available yet, there is a way for someone to make their own instance to receive the federated content. I think.
In comparison, the Sublinks project garnered so much attention that the Lemmy.World admins seemed ready to jump onto it the moment that it was halfway ready, but now we haven't heard of any progress whatsoever for months and people seem to have given up hope for it. So communication about a project can be quite important as well as the actual coding itself.
Still waiting for my account to be activated
The app is good, however what makes the other platforms good is their algorithm and the content that is available. If those are missing (and they likely will for privacy reasons) there's little reason for me to use it instead of instagram reels or tiktok for example.
not having it be littered by bigots and have an add every 4 scrolls, along with better privacy, is probably the reason why.
I mean... maybe? Those are nice things of course, but if I wanted to start using an alternative to TikTok I would never consider something without an algorithm. The whole point - for me at least - is to waste time eating digital slop served to you by an algorithm so that you don't have to know what you want to watch when you sit down and open it up. If I already knew what I wanted to watch I wouldn't be on an app like this, I'd be searching up the video or subject on another platform.
You don't know what you get though, even without an AI driven "engagement" algorithm. If the pace of new videos being uploaded is fast enough there will be no shortage of hours to mindlessly scroll (if that's your goal). It gives everyone an equal chance to be discovered, and from there more popular videos will have their own tabs so you can see what other people like too. The short video format is plenty addictive without pushing for maximum retention, because there's no reason to - other platforms do it to increase your screen time so they can show you more ads. That's all it is. All it ever was. They don't give a shit about you or your enjoyment(YouTube dislikes?), in fact, one method of raising engagement is to make you angry, they do this a lot. It works, it's a lesson they learned from the effectiveness of rage baits.
It's honestly frustrating how so many people want the same old shit algorithmiclly based feed when we've finally gotten rid of the fucking disease.
Is the Android app out yet?
Yes.
Hopefully they implant P2P-CDN , via webtorrent and webrtc to ease bandwidth costs.
I'm finally got it working, looks great! federation isn't there yet but I'll take it over the mastodon UI for video playback.